> SpecStream is the live web data engine for teams and agents. This is the Markdown copy of https://specstream.ai/integrations/amazon-bedrock-agentcore. Every page, with what SpecStream does and costs: https://specstream.ai/llms.txt

# Amazon Bedrock AgentCore integration

[Integrations](https://specstream.ai/integrations.md) / [Agent frameworks & builders](https://specstream.ai/integrations.md#frameworks)

# SpecStream for Amazon Bedrock AgentCore

SpecStream is the live web data engine for teams and agents. Add SpecStream as an MCP server target on an Amazon Bedrock AgentCore Gateway, and every agent behind the gateway can ask for live web data, priced before it runs, with each value's source.

[Read the guide](https://specstream.ai/docs/integrations/amazon-bedrock-agentcore.md)

Status

Live

Category

[Agent frameworks & builders](https://specstream.ai/integrations.md#frameworks)

Built by

SpecStream

You need

An AWS account with an AgentCore Gateway, and a SpecStream API key.

Pricing

[You pay for the matches](https://specstream.ai/pricing.md)

Setup guide

[Read the guide](https://specstream.ai/docs/integrations/amazon-bedrock-agentcore.md)

Help

[Talk to a human](https://specstream.ai/talk-to-a-human.md)

## What you get

-   **One target, every agent**Agents that call the gateway see SpecStream's tools as specstream\_\_\_compile, specstream\_\_\_quote\_capacity\_work, and the rest.
-   **The key stays in AWS**AgentCore Identity holds the SpecStream API key and adds it as a Bearer header on every call to the target.
-   **Limits on the key**Give the key a total spending limit and the operations it needs, and no agent behind the gateway spends past it.

## Set it up

1.  Create a SpecStream API key with a spending limit under Agents & API.
2.  Store it in AgentCore Identity as an API key credential provider.
3.  Add an MCP server target at https://api.specstream.ai/mcp with that key in the Authorization header.

[The full setup guide](https://specstream.ai/docs/integrations/amazon-bedrock-agentcore.md) has every step, with what to do if something doesn’t arrive.

## Related integrations

[See all](https://specstream.ai/integrations.md#frameworks)

-   [OpenAI APILive web records as an MCP tool in the Responses API](https://specstream.ai/integrations/openai.md)
-   [LangChainAgent tools and a retriever that returns LangChain Documents](https://specstream.ai/integrations/langchain.md)
-   [LlamaIndexA tool spec and retriever nodes for LlamaIndex agents](https://specstream.ai/integrations/llamaindex.md)
-   [CrewAILive web rows for the agents in your crew](https://specstream.ai/integrations/crewai.md)
-   [VercelAI SDK tools in TypeScript, plus a client for route handlers](https://specstream.ai/integrations/vercel.md)
-   [HaystackHaystack Documents from the live web, for pipelines and agents](https://specstream.ai/integrations/haystack.md)

## Live web data, in Amazon Bedrock AgentCore

Ask in your own words for companies, places, people, filings or anything else the web lists. You see the price before anything runs and pay per match.

[Read the guide](https://specstream.ai/docs/integrations/amazon-bedrock-agentcore.md)